154 / 220
Dual Lab Validation of a Method for the Determination of Fructans in Infant
Formula & Adult Nutritionals /
25 Jul 2016
44
With respect to sample 14, it should be noted that the low level fructan addition results in an very
small increase in the fructan concentration which is less than the LOQ value of the method.
In general it can be concluded that the established recoveries are within the SMPR target range.
f.
Specificity
In this specificity study different pure carbohydrates (disaccharides, oligosaccharides and starch) have
been subjected to the fructan analytical protocol. The aim of these analyzes is to assess whether, and
if so, to what extent, these non‐fructan carbohydrates contribute to falsely too high measured fructan
levels. To this end, a number of pure carbohydrate components with a test portion of 0.5 gram, has
been analyzed by using the fructan analysis method.
The analytical results are expressed in different ways.
Firstly
it is was calculated as a 4 gram liquid or reconstituted sample was analyzed. In that way the
above applied test portion of 0.5 gram pure carbohydrate components simulates the presence of this
constituent of 0.5/4 x 100% = 12.5 % level in a 4 gram liquid (reconstituted) test sample.
Secondly
the result was expressed as the fructan content in the dry product, thus in 100% investigated
matrix.
Moreover two different procedures have been followed for measuring and quantifying the fructan
content. The standard procedure, as given in the protocol, is the
“no A0 correction”
columns. In the
other procedure,
“with A0 correction”
, a blank subtraction step was introduced. In that case the
analytical protocol has been extended with a blanc measurement before incubating the processed
sample with the enzyme fructanase. This blanc measurement is done directly after the graphitized
carbon solid phase extraction (GC‐SPE) procedure. After the GC‐SPE the sample is split in two aliquots.
One aliquot is treated with the fructanase exactly conform the protocol and analyzed. In the other
aliquot no fructanase hydrolysis is carried out but is directly analyzed with the HPAEC‐PAD as a blank
measurement. Then final fructan content is calculated by subtracting the blank result from the result
obtained with the standard procedure. The difference is the amount of fructose and glucose which are
set free by the fructanase hydrolysis. In that way easily can be corrected amongst others for
chromatographic peaks which coincide with either fructose and glucose without being fructose or
glucose.
The results are presented in table 8.
Looking to the results in the column “expressed in liquid sample” it can be concluded that, with the
exception of Litesse II polydextrose, applying the standard protocol without blank correction (“no A0
correction”) of all potential interferences (at the high 12,5 % level in the sample to analyze!) the
erroneously measured fructan contents are less than the LOQ and therefore of no concern. For the
Litesse II polydextrose the erroneously measured fructan content is at the level of the LOQ.
In the table for the constituents Vivinal GOS and Litesse II polydextrose, the worst case scenario’s are
presented. The chromatograms (figure 1 and 2) of these two constituents shows chromatographic
signals near the fructose and glucose peaks with retention times which differs somewhat of the
calibration standards. However in case the GOS and polydextrose constituents are present (at these
high concentration levels!) in fructan containing samples, it is very likeable that these peaks cannot
been distinguished from the fructose and glucose peaks coming from the hydrolyzed fructans.
VALIDATION REPORT
FOR ERP USE ONLY
DO NOT DISTRIBUTE